Data-led scouting: a working methodology

Data-led scouting is often described as if a database could pick players. It cannot. What data does well is narrow a search: from thousands of footballers, it produces a shortlist of names whose measured behaviour matches what a club is looking for. The judgement still belongs to people who watch.

This guide sets out a working scouting methodology — how a role is defined in numbers, how filters build a shortlist, how that shortlist is watched and how the findings are communicated back to the people who make decisions.

Defining the role before searching

A scouting search should begin with a definition, expressed in behaviour rather than in a position. What does this role need to do with the ball, how far should the player progress it, how much defensive work is required and in which zones? Written down, those requirements become the filters that the search will use.

Defining the role first prevents the most common failure in data-led recruitment: searching for the best players in a position rather than for the players who fit the role the team actually needs. A club does not recruit a profile in the abstract; it recruits a solution to a specific problem.

Building a shortlist with filters

Filters translate the role definition into queries. Rate metrics — progression per possession, pressing intensity, receiving locations, passing risk — are set to ranges that describe the requirement. League strength is used as context, so that a performance in a weaker division is judged with an appropriate discount.

The output is a shortlist, not a ranking. Analysts deliberately keep the list short, because the next stage is human and does not scale. Twenty names can be watched carefully; two hundred cannot. The filters exist to make the watching realistic.

From shortlist to eye test

The shortlist is then watched, and this is where data hands over. Video shows the qualities no event log holds: the timing of a run, the quality of a first touch under pressure, the way a player reacts to a mistake, the personality visible in a tough match. A player can pass every filter and fail the eye test, and that ends the search for them.

The reverse also happens. A player can sit outside the filters on one dimension and still be the right answer, because the role, the system or the price changes the calculation. Data-led scouting is a conversation between the numbers and the footage, and neither partner is optional.

Communicating the findings

A scouting report has to travel. It goes to a sporting director, a head coach and sometimes a board, each with a different appetite for numbers. The analyst's job is to present the evidence clearly: what the role required, which dimensions the player met, which remain uncertain and what the footage added.

Honesty about uncertainty is what makes a report useful. A profile that says "meets the pressing requirement, unproven in the receiving patterns this system uses" gives the decision-maker something to act on. A profile that claims certainty the data cannot support simply transfers risk without warning.

Stages of a data-led scouting process

A repeatable process moves from definition to decision in defined steps.

• Define the role in measurable behavioural terms.
• Set filters that translate the definition into a query.
• Apply a league-strength adjustment to contextualise the data.
• Produce a short, watchable list rather than a long ranking.
• Watch footage to test qualities that the data cannot capture.
• Report the evidence and the remaining uncertainty together.

How a long list becomes a decision

Each stage of a scouting process narrows the field. The bars show the relative volume of candidates surviving each filter.

What each scouting stage adds

Data and video contribute different evidence to the same decision.

The analyst's currency is attention. Every report competes with a coach's time, so the value lies in selecting what matters rather than in exhaustively describing what happened.

Football analysis sits between data and judgement. The data supplies a pattern; the human supplies the context, the priorities and the decision that no dataset can make alone.

A workflow is a set of definitions repeated. Agreeing those definitions in advance is what allows results to be compared across matches, players and seasons.

Automation frees time for the parts of analysis that need a human: interpreting context, choosing what to show and communicating it clearly.

Data narrows, people decide

Data-led scouting is powerful precisely because it knows its limits. It compresses thousands of players into a list a scout can actually watch, and then it steps back. The value comes from using it that way — as a filter, never as a verdict.